aiagents4pharma 1.45.1__py3-none-any.whl → 1.46.1__py3-none-any.whl

aiagents4pharma/talk2knowledgegraphs/configs/utils/database/milvus/default.yaml

@@ -0,0 +1,61 @@
+# Milvus Database Configuration
+# This config is used by backend tools and the MilvusConnectionManager
+# Separated from frontend config for proper backend-frontend separation
+#
+# Environment Variables (all optional with sensible defaults):
+#   MILVUS_HOST - Milvus server host (default: localhost)
+#   MILVUS_PORT - Milvus server port (default: 19530)
+#   MILVUS_USER - Milvus username (default: root)
+#   MILVUS_PASSWORD - Milvus password (default: Milvus)
+#   MILVUS_DATABASE - Database name (default: t2kg_primekg)
+
+milvus_db:
+  # Connection settings
+  alias: "default"
+  host: ${oc.env:MILVUS_HOST,localhost}
+  port: ${oc.env:MILVUS_PORT,19530}
+  uri: "http://${oc.env:MILVUS_HOST,localhost}:${oc.env:MILVUS_PORT,19530}"
+  token: "${oc.env:MILVUS_USER,root}:${oc.env:MILVUS_PASSWORD,Milvus}"
+  user: ${oc.env:MILVUS_USER,root}
+  password: ${oc.env:MILVUS_PASSWORD,Milvus}
+
+  # Database and collection names
+  database_name: ${oc.env:MILVUS_DATABASE,t2kg_primekg}
+  collection_edges: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_edges"
+  collection_nodes: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes"
+  collection_nodes_gene_protein: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes_gene_protein"
+  collection_nodes_molecular_function: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes_molecular_function"
+  collection_nodes_cellular_component: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes_cellular_component"
+  collection_nodes_biological_process: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes_biological_process"
+  collection_nodes_drug: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes_drug"
+  collection_nodes_disease: "${oc.env:MILVUS_DATABASE,t2kg_primekg}_nodes_disease"
+
+  # Query performance settings
+  query_batch_size: 10000
+
+# Node and edge column mappings
+node_id_column: "node_id"
+node_attr_column: "node_attr"
+edge_src_column: "edge_src"
+edge_attr_column: "edge_attr"
+edge_dst_column: "edge_dst"
+
+# Node colors for visualization (moved from frontend)
+node_colors_dict:
+  "gene/protein": "#6a79f7"
+  "molecular_function": "#82cafc"
+  "cellular_component": "#3f9b0b"
+  "biological_process": "#c5c9c7"
+  "drug": "#c4a661"
+  "disease": "#80013f"
+
+# BioBridge data source configuration
+biobridge:
+  source: "/mnt/blockstorage/biobridge_multimodal/"
+  node_type:
+    - "gene/protein"
+    - "molecular_function"
+    - "cellular_component"
+    - "biological_process"
+    - "drug"
+    - "disease"

aiagents4pharma/talk2knowledgegraphs/entrypoint.sh

@@ -173,18 +173,8 @@ log "Data loading phase completed. Starting main application..."
 # Ensure Python path includes the app directory
 export PYTHONPATH="/app:${PYTHONPATH}"
 
-# Create cache directory and set path for container
-cache_dir="/app/aiagents4pharma/talk2knowledgegraphs/tests/files"
-if [ ! -d "$cache_dir" ]; then
-	log "Creating cache directory: $cache_dir"
-	mkdir -p "$cache_dir"
-fi
-
-# Set container-specific cache path
-export CACHE_EDGE_INDEX_PATH="/app/aiagents4pharma/talk2knowledgegraphs/tests/files/t2kg_primekg_edge_index.pkl"
-
 log "Starting main application..."
 log "Python path: $PYTHONPATH"
-log "Cache edge index path: $CACHE_EDGE_INDEX_PATH"
+log "Note: Edge index is now loaded on-demand from Milvus (no cache file needed)"
 log "Executing command: $@"
 exec "$@"

aiagents4pharma/talk2knowledgegraphs/milvus_data_dump.py

@@ -58,21 +58,22 @@ class SystemDetector:
         if self.use_gpu and self.os_type == "linux":
             # Exact package list from original script for GPU mode
             packages = [
-                # "pip install --extra-index-url=https://pypi.nvidia.com cudf-cu12",
-                # "pip install --extra-index-url=https://pypi.nvidia.com dask-cudf-cu12",
-                "pip install pymilvus==2.5.11",
-                "pip install numpy==1.26.4",
-                "pip install pandas==2.1.3",
-                "pip install tqdm==4.67.1",
+                "pip install --extra-index-url=https://pypi.nvidia.com cudf-cu12",
+                "pip install --extra-index-url=https://pypi.nvidia.com dask-cudf-cu12",
+                "pip install pymilvus",
+                "pip install numpy",
+                "pip install pandas",
+                "pip install tqdm",
             ]
             return packages
         else:
             # CPU-only packages
             packages = [
-                "pip install pymilvus==2.5.11",
-                "pip install numpy==1.26.4",
-                "pip install pandas==2.1.3",
-                "pip install tqdm==4.67.1",
+                "pip install pymilvus",
+                "pip install numpy",
+                "pip install pandas",
+                "pip install tqdm",
+                "pip install pyarrow",
             ]
             return packages
 

aiagents4pharma/talk2knowledgegraphs/tests/test_agents_t2kg_agent.py

@@ -2,14 +2,19 @@
 Test cases for agents/t2kg_agent.py
 """
 
+from contextlib import ExitStack
 from unittest.mock import MagicMock, patch
 
 import pandas as pd
 import pytest
-from langchain_core.messages import HumanMessage
+from langchain_core.messages import HumanMessage, ToolMessage
 from langchain_openai import ChatOpenAI, OpenAIEmbeddings
+from langgraph.types import Command
 
 from ..agents.t2kg_agent import get_app
+from ..tools.milvus_multimodal_subgraph_extraction import (
+    MultimodalSubgraphExtractionTool,
+)
 
 DATA_PATH = "aiagents4pharma/talk2knowledgegraphs/tests/files"
 
@@ -57,6 +62,8 @@ def mock_milvus_collection(name):
     """
     Mock Milvus collection for testing.
     """
+    # name is intentionally unused in this simplified mock
+    del name
     nodes = MagicMock()
     nodes.query.return_value = [
         {
@@ -98,88 +105,202 @@ def mock_milvus_collection(name):
     ]
     edges.load.return_value = None
 
-    if "nodes" in name:
-        return nodes
-    if "edges" in name:
-        return edges
-    return None
+    # Default path in tests expects None for unknown collections (implicit)
 
 
-def test_t2kg_agent_openai_milvus_mock(input_dict):
-    """
-    Test the T2KG agent using OpenAI model and Milvus mock.
-
-    Args:
-        input_dict: Input dictionary
-    """
-    input_dict["llm_model"] = ChatOpenAI(model="gpt-4o-mini", temperature=0.0)
-    input_dict["embedding_model"] = OpenAIEmbeddings(model="text-embedding-3-small")
-    unique_id = 12345
+def _invoke_app_with_mocks(unique_id, input_dict):
+    """Run the app with patched Milvus + tool stack and return (app, config, response)."""
     app = get_app(unique_id, llm_model=input_dict["llm_model"])
     config = {"configurable": {"thread_id": unique_id}}
     app.update_state(config, input_dict)
-    prompt = """
-    Adalimumab is a fully human monoclonal antibody (IgG1)
-    that specifically binds to tumor necrosis factor-alpha (TNF-α), a pro-inflammatory cytokine.
-
-    I would like to get evidence from the knowledge graph about the mechanism of actions related to
-    Adalimumab in treating inflammatory bowel disease
-    (IBD). Please follow these steps:
-    - Extract a subgraph from the PrimeKG that contains information about Adalimumab.
-    - Summarize the extracted subgraph.
-    - Reason about the mechanism of action of Adalimumab in treating IBD.
-
-    Please set the extraction name for the extraction process as `subkg_12345`.
-    """
+    prompt = (
+        "Adalimumab is a fully human monoclonal antibody (IgG1) that "
+        "specifically binds to tumor necrosis factor-alpha (TNF-α), a "
+        "pro-inflammatory cytokine.\n\n"
+        "I would like to get evidence from the knowledge graph about the "
+        "mechanism of actions related to Adalimumab in treating inflammatory "
+        "bowel disease (IBD). Please follow these steps:\n"
+        "- Extract a subgraph from the PrimeKG that contains information about "
+        "Adalimumab.\n- Summarize the extracted subgraph.\n"
+        "- Reason about the mechanism of action of Adalimumab in treating IBD.\n\n"
+        "Please set the extraction name for the extraction process as `subkg_"
+        f"{unique_id}`."
+    )
+
+    mocks = {
+        "pcst": MagicMock(),
+        "connections": MagicMock(),
+        "compose": MagicMock(),
+        "connections_manager": MagicMock(),
+        "db": MagicMock(),
+        "conn_mgr": MagicMock(),
+    }
+
+    with ExitStack() as stack:
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.tools."
+                "milvus_multimodal_subgraph_extraction.Collection",
+                side_effect=mock_milvus_collection,
+            )
+        )
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.tools."
+                "milvus_multimodal_subgraph_extraction.MultimodalPCSTPruning",
+                mocks["pcst"],
+            )
+        )
+        stack.enter_context(patch("pymilvus.connections", mocks["connections"]))
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.tools."
+                "milvus_multimodal_subgraph_extraction.hydra.initialize"
+            )
+        )
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.tools."
+                "milvus_multimodal_subgraph_extraction.hydra.compose",
+                mocks["compose"],
+            )
+        )
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.utils.database."
+                "milvus_connection_manager.connections",
+                mocks["connections_manager"],
+            )
+        )
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.utils.database."
+                "milvus_connection_manager.Collection",
+                side_effect=mock_milvus_collection,
+            )
+        )
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.utils.database.milvus_connection_manager.db",
+                mocks["db"],
+            )
+        )
+        stack.enter_context(
+            patch(
+                "aiagents4pharma.talk2knowledgegraphs.tools."
+                "milvus_multimodal_subgraph_extraction.MilvusConnectionManager",
+                mocks["conn_mgr"],
+            )
+        )
 
-    with (
-        patch(
-            "aiagents4pharma.talk2knowledgegraphs.tools."
-            "milvus_multimodal_subgraph_extraction.Collection",
-            side_effect=mock_milvus_collection,
-        ),
-        patch(
-            "aiagents4pharma.talk2knowledgegraphs.tools."
-            "milvus_multimodal_subgraph_extraction.MultimodalPCSTPruning"
-        ) as mock_pcst,
-        patch("pymilvus.connections") as mock_connections,
-        patch(
-            "aiagents4pharma.talk2knowledgegraphs.tools."
-            "milvus_multimodal_subgraph_extraction.hydra.initialize"
-        ),
-        patch(
-            "aiagents4pharma.talk2knowledgegraphs.tools."
-            "milvus_multimodal_subgraph_extraction.hydra.compose"
-        ) as mock_compose,
-    ):
-        mock_connections.has_connection.return_value = True
-        mock_pcst_instance = MagicMock()
-        mock_pcst_instance.extract_subgraph.return_value = {
+        def mock_tool_execution(tool_call_id, state, prompt, arg_data=None):
+            del prompt, arg_data
+            mock_extracted_graph = {
+                "name": f"subkg_{unique_id}",
+                "tool_call_id": tool_call_id,
+                "graph_source": "BioBridge",
+                "topk_nodes": 3,
+                "topk_edges": 3,
+                "graph_dict": {
+                    "name": "extracted_subgraph",
+                    "nodes": ["Adalimumab", "TNF"],
+                    "edges": [("Adalimumab", "acts_on", "TNF")],
+                },
+                "graph_text": (
+                    "Adalimumab acts on TNF protein for treating inflammatory diseases."
+                ),
+                "graph_summary": None,
+            }
+            tool_message = ToolMessage(
+                content=(
+                    "Subgraph extraction completed successfully. "
+                    "Extracted subgraph containing Adalimumab and TNF interactions."
+                ),
+                tool_call_id=tool_call_id,
+                name="subgraph_extraction",
+            )
+            return Command(
+                update={
+                    "messages": [tool_message],
+                    "dic_extracted_graph": state.get("dic_extracted_graph", [])
+                    + [mock_extracted_graph],
+                }
+            )
+
+        stack.enter_context(
+            patch.object(MultimodalSubgraphExtractionTool, "_run", side_effect=mock_tool_execution)
+        )
+
+        # set return values via the mocks dict
+        mocks["connections"].has_connection.return_value = True
+        mocks["connections_manager"].has_connection.return_value = True
+        mocks["db"].using_database.return_value = None
+
+        pcst_instance = MagicMock()
+        pcst_instance.extract_subgraph.return_value = {
             "nodes": pd.Series([0, 1]),
             "edges": pd.Series([0]),
         }
-        mock_pcst.return_value = mock_pcst_instance
-        mock_cfg = MagicMock()
-        mock_cfg.cost_e = 1.0
-        mock_cfg.c_const = 1.0
-        mock_cfg.root = 0
-        mock_cfg.num_clusters = 1
-        mock_cfg.pruning = True
-        mock_cfg.verbosity_level = 0
-        mock_cfg.search_metric_type = "L2"
-        mock_cfg.node_colors_dict = {"drug": "blue", "gene/protein": "red"}
-        mock_compose.return_value = MagicMock()
-        mock_compose.return_value.tools.multimodal_subgraph_extraction = mock_cfg
-        mock_compose.return_value.tools.subgraph_summarization.prompt_subgraph_summarization = (
+        mocks["pcst"].return_value = pcst_instance
+
+        cfg = MagicMock()
+        for k, v in {
+            "cost_e": 1.0,
+            "c_const": 1.0,
+            "root": 0,
+            "num_clusters": 1,
+            "pruning": True,
+            "verbosity_level": 0,
+            "search_metric_type": "L2",
+        }.items():
+            setattr(cfg, k, v)
+        cfg.node_colors_dict = {"drug": "blue", "gene/protein": "red"}
+
+        mocks["compose"].return_value = MagicMock()
+        mocks["compose"].return_value.tools.multimodal_subgraph_extraction = cfg
+        mocks[
+            "compose"
+        ].return_value.tools.subgraph_summarization.prompt_subgraph_summarization = (
             "Summarize the following subgraph: {textualized_subgraph}"
         )
 
+        db_cfg = MagicMock()
+        for k, v in {
+            "alias": "test_alias",
+            "host": "localhost",
+            "port": "19530",
+            "user": "root",
+            "password": "password",
+            "database_name": "test_db",
+        }.items():
+            setattr(db_cfg.milvus_db, k, v)
+        mocks["compose"].return_value.utils.database.milvus = db_cfg.milvus_db
+
+        conn = MagicMock()
+        conn.ensure_connection.return_value = True
+        conn.get_connection_info.return_value = {"database": "test_db", "connected": True}
+        conn.test_connection.return_value = True
+        mocks["conn_mgr"].return_value = conn
+
         response = app.invoke({"messages": [HumanMessage(content=prompt)]}, config=config)
 
-    assistant_msg = response["messages"][-1].content
-    assert isinstance(assistant_msg, str)
-    current_state = app.get_state(config)
-    dic_extracted_graph = current_state.values["dic_extracted_graph"][0]
+    return app, config, response
+
+
+def test_t2kg_agent_openai_milvus_mock(input_dict):
+    """
+    Test the T2KG agent using OpenAI model and Milvus mock.
+
+    Args:
+        input_dict: Input dictionary
+    """
+    input_dict["llm_model"] = ChatOpenAI(model="gpt-4o-mini", temperature=0.0)
+    input_dict["embedding_model"] = OpenAIEmbeddings(model="text-embedding-3-small")
+    unique_id = 12345
+    app, config, response = _invoke_app_with_mocks(unique_id, input_dict)
+
+    assert isinstance(response["messages"][-1].content, str)
+    dic_extracted_graph = app.get_state(config).values["dic_extracted_graph"][0]
     assert isinstance(dic_extracted_graph, dict)
     assert dic_extracted_graph["name"] == "subkg_12345"
     assert dic_extracted_graph["graph_source"] == "BioBridge"
@@ -190,9 +311,8 @@ def test_t2kg_agent_openai_milvus_mock(input_dict):
     assert len(dic_extracted_graph["graph_dict"]["edges"]) > 0
     assert isinstance(dic_extracted_graph["graph_text"], str)
     assert isinstance(dic_extracted_graph["graph_summary"], str)
-    assert "Adalimumab" in assistant_msg
-    assert "TNF" in assistant_msg
+    assert "Adalimumab" in response["messages"][-1].content
+    assert "TNF" in response["messages"][-1].content
 
     # Another test for unknown collection
-    result = mock_milvus_collection("unknown")
-    assert result is None
+    assert mock_milvus_collection("unknown") is None